A pet owner asks Ridgeline's service agent, 'Is Rimadyl 75mg still on the recall list?' Which search type is best suited to finding the exact drug name and dosage in the indexed policy documents?
Select an answer to reveal the explanation.
Short Explanation
When the question hinges on an exact string like a drug name and dosage, keyword search is the right tool — it's built to find precise matches, not paraphrased meaning.
Full Explanation
Keyword search matches exact terms and phrases, which makes it the strongest fit for queries built around specific, literal strings such as a drug name and dosage — 'Rimadyl 75mg' is exactly the kind of precise token that keyword matching is designed to find reliably. Vector search excels at semantic similarity and paraphrase, but claiming it 'always outperforms' keyword search overstates its strength; vector search can actually underperform on exact-match lookups like product names, codes, or dosages because it optimizes for meaning rather than literal string matching. Dismissing hybrid search as never appropriate for pharmacy queries is also wrong — hybrid combining keyword precision with vector recall can be a reasonable choice when queries mix exact terms with looser phrasing, it's just not strictly required for this narrow exact-match case. The claim that drug recalls exist only as picklist values is an unfounded assumption not supported by the scenario, which describes recall information living in indexed policy documents. The exam-relevant principle: choose keyword search when exact terminology matters, vector search when paraphrase and meaning matter, and hybrid when both matter.